itwiki18fandomcom-20200213-history
It Wiki
Welcome to the ! Deep Learning Deep Learning is a subcategory of machine learning, which uses algorithms and patterns in order for a machine to complete a task that it was not specifically asked to do. Essentially, the computer is using its own "intuition". It acts and works like a human brain especially due to the fact that it can add data or information in layers. This means the machine can "think" for itself. Deep learning has some promising qualities. Deep Learning does not require well-labeled data. Another positive quality is that it reduces the need for feature-engineering which is known for its lengthy process. Lastly this technology is easily adaptable when problems occur. Having listed some of the pros of deep learning, there are some negative qualities. Deep Learning requires a large number of data in order to function properly. This is mainly due to the fact that in order to pull out random data you have to have large quantities of that data. For example, imagine if a person were to look up the word "apple". There are many types of apples- granny smith, red gala, etc.- many of which have different colors or were grown in different places or countries. Possibly, the person may have not meant a fruit at all but instead was referring to the popular phone company. In order to account for the many combinations of the word "apple", the data must include the variations of the word as well as information about the word. Another negative quality is deep learning is an expensive project to oversee, there is no specific guide on how to learn how to construct deep learning, and what is learned is not easily understood. This field requires highly skilled and trained professionals. Another negative quality of deep learning is how a person might use this technology. One of the scariest ways a person could misuse deep learning is described by a concept called deep fake. This technology can damage reputations by allowing the user to create an event, specifically of a person, that the said person never did. For example, the user could record themselves saying an offensive phrase, then match their audio and video to another individual to make it seem as if that person had said the offensive phrase. This concept may seem far-fetched but imagine if the CEO of a major company was recorded saying an offensive phrase that was posted on social media. The CEO will not only give his company a bad reputation but he would most likely also lose his job. Important articles File:Placeholder | A page about your topic |link=A page about your topic File:Placeholder | A Main Character |link=A Main Character File:Placeholder | The First Episode |link=The First Episode File:Placeholder | An Important Location |link=An Important Location File:Placeholder | A Key Event |link=A Key Event File:Placeholder | A Crucial Item |link=A Crucial Item Need help building out this community? * * * * * You can also be part of the larger Fandom family of communities. Visit Fandom's Community Central! Category:Browse